Raman Spectrum Matching with Contrastive Representation Learning

TL;DR

This paper introduces a contrastive learning approach for Raman spectrum matching that requires minimal preprocessing and training data, achieving competitive accuracy and enabling confidence set computation.

Contribution

It presents a novel contrastive representation learning method for Raman spectra that outperforms or matches existing techniques with less data and preprocessing.

Findings

Significantly improves prediction accuracy on three datasets

Works effectively with as little as one reference spectrum per class

Enables computation of conformal prediction sets with guaranteed coverage

Abstract

Raman spectroscopy is an effective, low-cost, non-intrusive technique often used for chemical identification. Typical approaches are based on matching observations to a reference database, which requires careful preprocessing, or supervised machine learning, which requires a fairly large number of training observations from each class. We propose a new machine learning technique for Raman spectrum matching, based on contrastive representation learning, that requires no preprocessing and works with as little as a single reference spectrum from each class. On three datasets we demonstrate that our approach significantly improves or is on par with the state of the art in prediction accuracy, and we show how to compute conformal prediction sets with specified frequentist coverage. Based on our findings, we believe contrastive representation learning is a promising alternative to existing methods for Raman spectrum matching.